Many industrial noise environments are characterized by high levels of impact noise which are invariably superimposed on a continuous background noise which itself is very "peaked" (high kurtosis), i.e., has a non-Gaussian distribution of amplitudes. The limited demographic data available show that such environments pose an unusually high risk of hearing loss with prolonged or repeated exposures. Our laboratory results from animals exposed to idealized combinations of continuous and impact noise agree with the demographic data in showing an exacerbation of hearing loss following exposures to such a noise combination. There is a general consensus that we do not, as yet, have an adequate descriptor or measure of complex acoustic environments which can be used for the purpose of evaluating the risk to hearing and establishing exposure standards. The hearing hazards from complex noise environments will be studied using the chinchilla as an animal model. Industrial noise environments will be modeled in two totally different ways: (1) by producing idealized combinations of a specified band on continuous noise to which a discrete series of impact noises is added. By studying the effects of parameter variations on the hearing of experimental animals we can learn something of the limits of the interaction between these two classes of noise. While we are currently pursuing this approach, and propose a limited extension of these studies, the approach has practical limitations. (2) by digitally generating a continuous noise within which the impulsiveness (kurtosis) can be closely controlled. We will also be able to control from which part of the audible spectrum the energy for the impulsive peaks arises. This approach will yield the most realistic modeling and holds the greatest promise for producing an index by which to gauge the hazards posed by virtually any industrial noise environment. We propose to synthesize such noise and expose groups of animals to these synthesized noise conditions. The kurtosis (B2), spectral variables and level will be systemically varied. Hearing trauma will be assessed in terms of audibility changes (thresholds and tuning curves) measured using the evoked auditory response, and sensory cell losses in the cochlea. We will test the hypothesis that a high B2 noise exposure is more hazardous to hearing than is a low B2 exposure, of the same energy and that this effect is frequency dependent. The truth of this statement will allow us to develop a physical measure of a complex noise exposure, (which correlates with our indices of hearing trauma) for the purpose of evaluating hazards to hearing.